Page:EB1922 - Volume 30.djvu/65

(450 C.), and later the tempering was abandoned. Round swaged tie-rods were made from the same steel as the streamline wires drawn to an ultimate strength of 80 tons per sq. in. without any sub- sequent heat treatment. The adoption of tie-rods and streamline wires for bracing extended the period for which the aeroplane retained its truth, while it was improved both in speed and climb by the fair wires.

Flexible cables used for controls consisted generally of seven strands, each of 19 wires of go-ton tensile. To increase the war out-

Eut a single lay cable of larger strands was used. This cable could not e spliced, and joints were made by turning the ends, wrapping with wire, and soldering a process that requires much care.

In a few cases in 1919 the structures were built on the strut-tie principle without wire bracing; this gave quick erection and main- tained very well the truth of structure.

Dope. The fabric stretched over the wings becomes slack after exposure to alternations of humidity. Prior to 1909 rubber cotton fabric was tried, and alternatively the plain cotton was tautened by painting with flour paste. In 1909 thin sheets of cellulose acetate were applied over the cotton, and later the substance was dissolved in acetone and applied with a brush, camphor being used to keep the coat pliable ; however, the camphor evaporated, and thereupon the dope cracked on exposure. The search for a suitable softener that did not evaporate from the dope was prosecuted. Tetrachlorethane was tried with success, but it proved dangerous to the operatives applying it in enclosed places. Moreover, sunlight decomposed tetrachlorethane; to yield hydrochloric acid, which eventually at- tacked the fabric.

In 1916 benzyl-alcohol was tried with success. When the evil effect of light on linen and dope was discovered in England a pig- ment varnish was introduced by the Royal Aircraft Factory (P.C.io) which protected the fabric and dope from light and increased the life of both. In 1916 a nitrocellulose dope was introduced, to economize in the acetic-acid radicals which were in demand elsewhere for ex- plosives. From 1916 onwards the acetate and nitro dopes were used; to them benzyl-alcohol was added to render the film plastic, and triphenyl-phosphate to render the film waterproof and fireproof. After removing all saponifiable grease from the fabric the dope was applied by hand in three to five coats, till 1918, when spraying was introduced for the coats other than the first, which needed to be well brushed in. Constant temperature and low humidity are required in dope-rooms to avoid the deposit of water due to evaporation of the solvents.

Rubber Hose. Rubber tube introduced in pipe lines to give flexi- bility is deteriorated by petrol and oil. In 1916 some resistance to petrol was introduced by using pure para heavily loaded with mineral matter and rather over-vulcanized. This withstood boiling petrol for one hour, and immersion for 23 hours, but its life in use is very short and it frequently required renewal after four months.

Engine Testing. The airscrew, the flat plate air brake, the electric dynamo and the water dynamometer of Heenan and Froude were used for testing aero engines.

Later the Escargot reaction torque brake was evolved, correspond- ing in principle to the Heenan and Froude water brake in that an air fan brake is rotated inside a closely fitted case, into which the air is drawn through central ports and expelled centrifugally through tangential outlets at top and bottom; the engine, mounted on a

FIG. 19.

built-up stand pivoted about the propeller shaft axis, is held and the torque measured with a graduated bar and counterpoise.

To vary the power absorbed at a given speed, the Fell type of Escargot (see figs. 19 and iga), introducing Butterfly valves in the tangential outlets, was developed late in 1917. Restriction of the air outlet from the Escargot perforce reduces the work to be done by the fan on the air, which tends to rotate with the fan and so increase the speed of the engine to a corresponding degree. A power curve range is thus obtained comparable with that given by the Heenan and Froude brake. The Escargot method provides a ready means of cooling air-cooled engines by taking special ducts from the outlets to the engine cylinders, whereas the Heenan and Froude brake re- quires a separate cooling-fan and driving-motor.

In determining the useful H.P. of rotary engines, " windage loss," or the power absorbed by the engine itself, had first to be determined for each type, and then deducted from the total nominal power, calculated on the weight bar reading. Originally the bench tests comprised an endurance test of four hours, followed by complete dismantlement and examination for defective parts, excessive wear, reassembly, and final one-hour test, the engine being run throughout at normal speed and at full throttle, petrol and oil con- sumptions being recorded in both tests. Subsequently all-round experience and increased reliability of materials and their treatment permitted of a reduction of this endurance test, first to three hours, and then to two hours, with a final half-hour test. The engine through- out, save for the last five minutes, was throttled down to 90 % and sometimes even to 80 % of full power at normal speed, to prevent the overheating of and detonation in the relatively high compression engines. Such engines were designed to give full power at 5,000 ft. height rather than at ground level.

Standardization of the actual flow measurement of carburetter jets in place of orifice-diameter calibration made it possible to tune up engines for bench tests on a few minutes' running only. Also standard jets suitable for flight purposes were substituted for bench- test jets before delivery, so that the time of tuning-up on installation of the engine into the aircraft was diminished. In 1916 a petrol- flow meter, whereby the actual flow into each carburetter is indicated, facilitated the determination of petrol consumptions.

Crank- shafts. The aero-engine materials were covered by definite specifications originally issued by the Royal Aircraft Factory. The chemical composition was closely defined, heat treatment provided for, and an Izod Impact Test added to the usual tensile test, to give some indication as to the shock-resisting power of the material. The Izod Impact Test, though it does not reproduce the alternating and fatiguing stresses of actual running, has proved to be indispen- sable in detecting steel which has been improperly heat-treated. " Temper brittleness " induced in alloy steels by slow cooling from the tempering temperature, even after correct initial heat treatment, is detected by the 2-3 ft. Ib. obtained as compared with the 25-30 ft. Ib. with the identical steel if properly heat-treated and quenched after tempering. This brittleness, which obviously unfits steel for crank-shaft use, cannot be detected otherwise than by the impact test, since the usual tensile results and micro-structure examination in no way differentiate between the sound and temper-brittle conditions.

Early in 1915 the British Aeronautical Inspection directorate suggested the following nickel chromium steel for crank-shafts, connect-